Pauling.AI: Autonomous Computational Drug Discovery
Add an AI chemist to your team and accelerate your drug discovery program
Pauling.AI is an autonomous computational drug discovery platform. It gives your team world-class computational chemistry without hiring an in-house group or waiting on slow, expensive CROs.
Auralis Biotech represents Pauling.AI in the EU as its fractional commercial partner. We work as part of the Pauling.AI team in Europe, from the first conversation about your target to the delivery of the results.
Developed by Pauling.AI, Boston, USA. Represented in the EU by Auralis Biotech.
A ligand in a protein binding pocket. Image: Pauling.AI.
How it works
- Bring one targetStart with a scoped pilot: send us the protein target, structure or sequence, and the assay context you care about.
- We scope it with youTogether with the Pauling.AI scientists we define the screen and estimate the compute cost and the timeline before anything runs.
- The agent runs the workflowStructure preparation, pocket detection and screening of millions of compounds on cloud infrastructure, with automated quality control.
- Prioritization by simulationThe best hits are re-ranked with molecular dynamics and binding free energy estimates, to decide what goes to the lab.
- Ranked molecules for the labYou receive a ranked list of molecules with docking, molecular dynamics and free energy data, ready for wet-lab follow-up.
A fraction of the cost of a CRO project
Price is one of the main reasons teams choose Pauling.AI. There are no licence fees, tokens or per-seat costs, and no infrastructure to buy or maintain. Before your pilot starts, we estimate the compute cost and the timeline of your screen, so the budget is clear from day one.
Pauling.AI
- Cost
- A fraction of a CRO project. No licences, tokens or per-seat fees; compute cost estimated before you start.
- Time to results
- Days
- What you need
- A target and an assay context. No in-house computational chemistry team.
CRO project
- Cost
- Paid per project, with contracting overhead for every iteration.
- Time to results
- Weeks to months, including contracting and reports
- What you need
- Detailed specifications, and less control over the workflow
In-house pipeline
- Cost
- Software licences, GPU infrastructure and specialist staff.
- Time to results
- Months to build and integrate the pipeline
- What you need
- Computational chemists and IT support
Why biotechs use Pauling.AI
- No infrastructure to build. No GPU cluster, licence server, scripting or file-format wrangling.
- No CRO queue. Run and iterate in days instead of waiting weeks for contracting and reports.
- Physics-based prioritization. The AI orchestrates the workflow, while Uni-Dock, GROMACS and other state-of-the-art tools do the chemistry.
- Built for hard targets. Useful for selectivity, membrane proteins, ion channels, transporters, mutants and targets where docking alone is not enough.
What Pauling.AI runs today
Virtual screening
Large-scale virtual screening against protein targets, with docking against more than 10 million commercially available compounds or your own SDF library.
Target preparation
Receptor preparation, protonation, energy minimization and structure quality control, with binding pocket detection and interactive pocket selection.
Selectivity counter-screening
Screening against a second pocket or an off-target protein to prioritize selective molecules.
Molecular dynamics and ΔG
Molecular dynamics of the top hits with GROMACS, and MM-PBSA / ΔG ranking from the trajectories.
ADMET prediction
More than 80 absorption, distribution, metabolism, excretion and toxicity properties predicted for the candidates.
Reports
Pose quality control, 2D and 3D visualization, trajectories and ranked hit reports.
Customers and case studies
Used by biotech companies
Half of the best compounds with 7.7% of the docking
In a combinatorial space of 10,704 docked compounds, a synthon screening funnel on Pauling.AI recovered 53 of the true top 100 compounds while docking only 7.7% of the space: 6.7 times more than random selection on the same budget.
Selective ligands for JAK2 V617F
A counter-screening workflow docked the same ligands into the JH2 pocket of mutant JAK2 V617F and of wild-type JAK2, ranked them by selectivity and checked pose quality before molecular dynamics and wet-lab validation.
Molecular dynamics on membrane proteins
Pauling.AI runs molecular dynamics simulations for membrane protein targets, one of the target classes where docking alone is not enough.
Who it is for
- Academic groups and university spin-offs with a target and no computational chemistry team
- Biotech companies that want in silico candidates before they commit laboratory budget
- Medicinal chemists who want to run a campaign without building the pipeline themselves
Other technologies we support
Frequently asked questions
What is Pauling.AI?
Pauling.AI is an autonomous computational drug discovery platform. You describe your target, and its AI agent prepares the protein structure, identifies binding pockets, screens millions of compounds, re-scores the best hits with molecular dynamics, estimates binding free energy and returns a ranked list of molecules for wet-lab follow-up.
Do I need computational chemistry expertise to use it?
No. You provide the protein target, structure or sequence and the assay context, and the AI agent selects the protocols and configures the workflow.
Which methods does the platform use?
Docking with Uni-Dock against more than 10 million commercially available compounds or your own library, AI models such as Boltz and DiffDock, molecular dynamics with GROMACS, MM-PBSA / ΔG ranking, selectivity counter-screening and ADMET prediction.
How much does it cost?
A fraction of the cost of a CRO project, with no licence fees, tokens or per-seat costs. Before you start, we estimate the compute cost and the timeline of your screen together with the Pauling.AI team.
How long does a campaign take?
Days instead of months. The time depends on the target and on the size of the campaign.
How do I start?
With a scoped pilot on one target. Send us the target and the assay context, and we define the screen, the compute cost and the timeline together with the Pauling.AI team.
What is the role of Auralis Biotech?
Auralis Biotech represents Pauling.AI in the EU as its fractional commercial partner and acts as part of the Pauling.AI team in Europe.
Have a target in mind?
Bring us one target and the assay context you care about. We scope the pilot with the Pauling.AI team, estimate cost and timeline, and reply within 2 working days.